Grasping extreme aerodynamics on a low-dimensional manifold

Abstract

Modern air vehicles perform a wide range of operations, including transportation, defense, surveillance, and rescue. These aircraft can fly in calm conditions but avoid operations in gusty environments, encountered in urban canyons, over mountainous terrains, and in ship wakes. With extreme weather becoming ever more frequent due to global warming, it is anticipated that aircraft, especially those that are smaller in size, will encounter sizeable atmospheric disturbances and still be expected to achieve stable flight. However, there exists virtually no theoretical fluid-dynamic foundation to describe the influence of extreme vortical gusts on wings. To compound this difficulty, there is a large parameter space for gust-wing interactions. While such interactions are seemingly complex and different for each combination of gust parameters, we show that the fundamental physics behind extreme aerodynamics is far simpler and lower-rank than traditionally expected. We reveal that the nonlinear vortical flow field over time and parameter space can be compressed to only three variables with a lift-augmented autoencoder while holding the essence of the original high-dimensional physics. Extreme aerodynamic flows can be compressed through machine learning into a low-dimensional manifold, which can enable real-time sparse reconstruction, dynamical modeling, and control of extremely unsteady gusty flows. The present findings offer support for the stable flight of next-generation small air vehicles in atmosphere conditions traditionally considered unflyable.

Super-resolution analysis via machine learning: a survey for fluid flows

“…5. By visualizing the weight distribution between the latent space representation and the whole field, the shallow decoder provides nonlinear modes that represent the contribution of each latent variable for super-resolution reconstruction, which are analogous to those captured by nonlinear autoencoders [108,[115][116][117][118][119][120].…”

A cyclic perspective on transient gust encounters through the lens of persistent homology

“…Second, we use an autoencoder to transform the full-state dynamics of each gust encounter to a reduced-order space. The autoencoder is a data-driven method of feature extraction, capable of reducing complex fluid flows to a small number of essential state variables (Murata, Fukami & Fukagata 2020;Fukami & Taira 2023). In its basic configuration, the autoencoder is constructed as an approximation of an invertible, nonlinear transformation between a high-dimensional space and a low-dimensional space.…”

Cluster Globally, Model Locally: Clusterwise modeling of nonlinear dynamics

“…Many types of autoencoders [21][22][23][24][25] have shown good performance in capturing intricate and nonlinear dynamics. Some autoencoders have been designed to learn local mappings of nonlinear dynamics [26][27][28].…”